Nearly 1 out of every 10 adolescents in the U.S. suffers from severe obesity (age- and sex-specific body mass index (BMI) ?1.2 times the 95th percentile or ?35 kg/m2), a chronic and debilitating disease. Current treatment guidelines call for 3 to 6 months of lifestyle modification therapy (LSMT) followed by consideration of adjunct pharmacotherapy for patients failing to respond to LSMT alone. However, this arbitrary time recommendation to wait 3 to 6 months before initiating medication is not evidence based. Waiting 6 months before making a change may be too long for patients who are struggling to achieve BMI reduction. Yet, 3 months may not be long enough to assess patient responsiveness to LSMT and could therefor lead to over medication. Further, no guidance exists regarding how to manage patients who are non-responsive to first-line pharmacotherapy. For adults with obesity, treatment with pharmacotherapy routinely includes combination medications, such as phentermine+topiramate. Yet for adolescents with severe obesity, a group for whom limiting unnecessary medication exposure is highly valued, switching from one monotherapy to another with a different mechanism of action (e.g. switching from phentermine to topiramate) may be preferable to adding a second medication (e.g. phentermine+topiramate). This gap in evidence-based treatment guidelines that address the optimal timing and sequence of adjunct pharmacotherapy is difficult to fill because of the response heterogeneity to LSMT and pharmacotherapy; using a ?one-size-fits-all? approach will fail to help the non-responders and may unnecessarily ?over-treat? the responders. Instead, adaptive interventions have the potential to maximize outcomes for more patients while limiting risk from exposure to ineffective and/or needless medications. Specifically, adaptive interventions tailor therapy according to predictors of response and then adjust the therapy over time based on on-going performance. Thus, the overarching goal of this clinical trial is to garner data that will inform the development of an adaptive medical intervention for the treatment of severe adolescent obesity that includes empirically-derived decision rules that address when adjunct pharmacotherapy should be initiated, and if starting pharmacotherapy, what is the best sequence or course of therapy. This study will utilize a sequential multiple assignment randomized trial (SMART) methodology which was developed specifically for constructing adaptive interventions. One hundred fifty adolescents (age 12-18 years) with severe obesity will be recruited in this 2-staged SMART that will systematically examine: 1) the effect of a 3-month versus 6-month response assessment to LSMT before starting adjunct pharmacotherapy; and 2) for non-responders to initial adjunct phentermine, the relative effectiveness of adding topiramate to phentermine versus switching to topiramate monotherapy. All participants will receive 12 months of intervention and the primary outcome will be change in BMI at 12 months.